Disparities of Infant and Neonatal Mortality Trends In

Open access Research BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from Disparities of infant and neonatal mortality trends in Greece during the years of economic crisis by ethnicity, place of residence and human development index: a nationwide population study

Tania Siahanidou, 1 Nick Dessypris,2 Antonis Analitis,2 Constantinos Mihas,2 Evangelos Evangelou,3,4 George Chrousos,1 Eleni Petridou2,5

To cite: Siahanidou T, Abstract Strengths and limitations of this study Dessypris N, Analitis A, et al. Objective To study trends of infant mortality rate (IMR) Disparities of infant and and neonatal mortality rate in Greece during the period neonatal mortality trends in ►► Longest follow-up period of national data, as con- 2004–2016 and explore the role of sociodemographic Greece during the years of trasted to previously Greek-based studies. factors in the years of crisis. economic crisis by ethnicity, ►► Maternal ethnicity is for the first time considered in place of residence and human Design Nationwide individual data for live births and the analyses explaining part, if not most, of the het- development index: a nationwide infant (0–11 months) deaths provided by the Hellenic erogeneity of trends through the study period. population study. Statistical Authority were examined using Poisson, BMJ Open ►► Three complimentary statistical methodologies were 2019;9:e025287. doi:10.1136/ joinpoint regression and interrupted time series (ITS) used to describe the overall evolution of infant mor- bmjopen-2018-025287 analyses. tality and its components. Setting Greece. ►► Prepublication history and ►► Absence of a linkage system between birth and additional material for this Participants All infant deaths (n=4862) over the 13-year death data. paper are available online. To period, of which 87.2% were born to Greek mothers, and ►► Data on the role of biological factors or specific view these files, please visit respective live births.

causes of infant death are missing. http://bmjopen.bmj.com/ the journal online (http://dx.doi. Main outcome measures Evolution of IMR (0–364 days), org/10. 1136/bmjopen- 2018- early (<7 days) neonatal mortality rate (ENMR), late (7–27 025287). days) neonatal mortality rate (LNMR) and post neonatal IMR, LNMR and PNMR after 2012, and in ENMR after 2010, Received 7 July 2018 (28–364 days) mortality rate (PNMR) trends, by maternal among Greek mothers’ infants. Revised 30 May 2019 nationality, place of residence and Human Development Conclusions HDI and rural residence were significantly Accepted 2 July 2019 Index (HDI). associated with IMR. The strongly decreasing IMR trends Results By Poisson regression, overall, during the study among Greek-mothers’ infants were stagnated after a lag period, among infants of Greek mothers, IMR and PNMR time of ~4 years of crisis approximating the previously

declined significantly (−0.9%; 95% CI −1.7% to −0.1% sharply increasing trends among non-Greeks. on October 2, 2021 by guest. Protected copyright. and −1.6%; −3.0% to −0.2% annually, respectively),

although differentially by place of residence (IMRurban:

−2.1%; −2.9% to −1.3%, IMRrural: +10.6%; 7.6% to Introduction 13.6%). By contrast, among infants of non-Greek mothers, Socioeconomic factors have repeatedly been the low starting IMR/ENMR/LNMR/PNMR increased recognised as strong determinants of the significantly (max ENMR:+12.5%; 8.6% to 16.5%) health status of the population, including leading to a non-significant time–trend pattern overall in infants.1–3 Economic indicators, such as © Author(s) (or their Greece. The inverse associations of HDI with IMR, ENMR the per capita gross national product, and PNMR were restricted to Greek mothers’ infants. employer(s)) 2019. Re-use were suggested to be at least as important permitted under CC BY-NC. No Joinpoint regression analyses among Greek mothers’ contributors of infant mortality (IM) as commercial re-use. See rights infants indicated non-significant increasing trends of and permissions. Published by IMR and ENMR following the crisis (+9.3%, 2012–2016, narrowly defined factors relating to provi- BMJ. p=0.07 and +10.2%, 2011–2016, p=0.06, respectively). sion of medical care, for example, the rela- For numbered affiliations see By contrast, the high (+17.1%; 8.1% to 26.9%, p=0.002) tive number of doctors or hospital beds in end of article. 1 IMR increases among non-Greek infants were restricted a community. Not surprisingly, the recent Correspondence to to 2004–2011 and equalised to those of Greek mothers’ economic crisis of ~2008, has been linked Professor Tania Siahanidou; infants thereafter. ITS analyses in preset years (2008, with declines in population and child health siahan@ med.uoa. gr 2010, 2012) identified significantly increasing trends in reflected also by increased IM rates (IMR).4 5

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 1 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from It is worth noting that in European Union countries, even period as a proxy of individual and collective measures of a minor 1% cut in government healthcare spending socioeconomic impact and healthcare access. was associated with significant increases in all mortality metrics, including neonatal and post neonatal IM.6 On the contrary, associations between the crisis and Materials and methods increased mortality have been questioned in other studies Following personal contact and a signed agreement, indi- showing that most indicators of population health, apart vidual data for all live births and infant (0–11 months) from those relevant to suicides and mental health, deaths were provided by the Hellenic Statistical Authority continued improving after crisis initiation.7–10 Moreover, (ELSTAT) for a 13-year period (2004–2016) in two sepa- the economic crisis has been associated with some bene- rate files: one including live births, and a second one, ficial effects, that is, decline in rates of road traffic acci- infant deaths; linkage of the two files was not possible as dents, and smoking cessation.4 9 Reasonably, the impact of the personal identification number was not available in the crisis on health depends on several factors including ELSTAT. Information on maternal demographic charac- the duration and intensity of the recession, the level of teristics, such as nationality and place of residence, as well healthcare achieved prior to the recession and the type as on the infants’ age at death, was also provided. of austerity measures applied, but also on the type of Mortality rates were calculated for infant (ΙΜ), early the population studied with the most vulnerable groups neonatal (0–6 days) (ENM), late neonatal (7–27 days) being disproportionally affected.4 11 12 Due to the latter, it (LNM) and post neonatal (28–364 days) period (PNM) was proposed that studies should focus on analysing sepa- using respective numbers of deaths over the number of rately the subgroups most influenced by the crisis instead live births per year. Annual percent of change (APC) of presenting results as averages in a population.4 during the study period was initially estimated through Greece has been markedly affected and still suffering univariate Poisson regression analysis using the under- the recent economic crisis. Between 2008 and 2016, lying population of each set as an offset variable. Subse- the country’s gross domestic product (GDP) per capita quently, data were stratified and analysed by maternal dropped by 26%, unemployment rate increased by more nationality (Greek vs non-Greek) and place of residence than 200% (from 7.6% to 23.3%) and the median dispos- (urban/semiurban vs rural). Joinpoint regression anal- able income decreased by 35%.13 14 The Gini index, which ysis22 was thereafter applied to automatically derive, by a measures income inequality, increased by 22% and the software programme, different segments in the mortality proportion of individuals at risk of poverty or social exclu- evolution curves overall as well as those by maternal sion climbed to almost 36% in 2016. Total health expen- nationality and place of residence. In details, the Join- ditures and government expenditure on healthcare were point regression analysis is applied to study varying trends both decreased by 34% per capita over the period 2008– over time in order to identify the time point(s) in which 2016,13 14 whereas unmet healthcare needs increased.15 the trend significantly changes. The location of the join- http://bmjopen.bmj.com/ Besides, during the last years, Greece experienced large point is not known a priori and is to be estimated from the refugee flows16 17 with an anticipated negative impact on data. Therefore, the software takes trend data (eg, IMR) the country's economy and population health indicators. and fits the simplest joinpoint model based on the data. Adverse effects of financial crisis on perinatal factors, The user supplies the minimum and maximum number such as low birth weight, preterm birth and stillbirth of joinpoints. The algorithm starts with the minimum rates, have been also observed18 19; however, sparse data number of joinpoint (eg, 0 joinpoints, which is a straight exist regarding the potential association of the crisis with line) and tests whether more joinpoints are statistically IMR. It has been initially reported that during the period significant and must be added to the model (up to that on October 2, 2021 by guest. Protected copyright. 2003–2012, IMR in Greece did not differ between the maximum number). Significance is tested using a Monte precrisis and crisis period whereas a later study showed Carlo Permutation method. In all models examined in that IMR increased between 2010 and 2015 as contrasted this study, a minimum number of 0 joinpoints and a to the steady decrease observed during the preceding maximum 1 of 2 joinpoints were tested. Interrupted time decade.20 21 Ecological correlations may be prone, series (ITS) analyses were also undertaken as a sensitivity however, to fallacies, whereas the different parts shaping analysis to further explore the effect of crisis on IMR and the two main components of IM, namely neonatal (NM) its components in alternative prespecified years, notably and post neonatal mortality (PNM), should have been 2008, considered to indicate the initiation of the crisis distinctly examined along with other factors possibly in Greece—as the value of the GDP of the country was influencing the infants’ health, especially socioeconomic maximum at that year but dropped afterwards13—and status and access to healthcare delivery. in two successive periods 2 years apart (2010 and 2012), The aims of the current study were to explore time to control for possible time lags in observing the impact trends in early (<7 days of life) and late (7–27 days) NM, of the crisis. All ITS models were verified and appropri- PNM (28–364 days) and total IM (0–364 days) in Greece ately adjusted for possible autocorrelation; autoregressive during the period 2004–2016 after taking into account integrated moving average models were used to control nationality and place of residence, as well as changes in for autocorrelation and to estimate treatment effects the Human Development Index (HDI) during the study over multiple periods. In our ITS analysis, a slope ‘

2 Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from of interest’ is fitted until the introduction of the crisis, Results ‘>year of interest’ represents the change in the level of The annual average number of live births during the the mortality immediately following the initiation of the study period was 105 077, ranging from a high 118 302 crisis and ‘change of slope’ represents the differences (2008) to a low 91 847 live births in 2015. The annual between precrisis and postcrisis intervention slopes. average proportion of Greek mothers was 83.7%, whereas HDI is a summary measure of average achievements the remaining were mainly economic migrants. During of a country’s population in three areas including life the period 2004–2016, a total of 4862 infant deaths were expectancy, education and per capita income indicators. recorded of whom 4238 (87.2%) were of infants born to Annual HDI values for the underlying populations in the Greek mothers, whereas among the remaining 615 deaths year of death were extracted from the United Nations of infants born to non-Greek mothers, 298 (48.5%) Development Programme website (http://hdr .undp. were of infants born to Albanian mothers, 146 (23.7%) org/en/composite/HDI), whereas Poisson regression of infants born to mothers from Balkan countries and analysis was used to explore the association of 1 SD of countries of the former Soviet Union, 77 (12.5%) from HDI with total IM and its components. All statistical anal- Asia, 47 (7.6%) from countries of the European Union yses were performed using the SAS software (V.9.4, SAS or other developed countries and 47 (7.6%) from Africa. Institute), Stata software programme, V.13 (StataCorp. During the early and late neonatal period, 2107 (43.3%) 2013. Stata Statistical Software: Release 13. College and 1136 (23.4%) deaths occurred, respectively, whereas Station, Texas, USA: StataCorp LP) and Joinpoint Regres- another 1617 (33.3%) of deaths were recorded in the sion Program (Joinpoint Regression Program, V.4.5.0.1, post neonatal period. The annual average IMR was 3.5 June 2017; Statistical Methodology and Applications over 1000 live births, whereas the corresponding figures Branch, Surveillance Research Programme, National for the early, late and post neonatal period were 1.5‰, Cancer Institute). 0.8‰ and 1.2‰, respectively. The number of live births and infant deaths, total IMR per 1000 live births in infants Patient involvement born to Greek and non-Greek mothers as well as HDI No patients were involved in setting the research ques- values in Greece during the study period are shown in tion or the outcome measures, nor were they involved table 1. Moreover, infant deaths and total IMR by place in developing plans for design or implementation of the of residence (urban/semiurban vs rural) are shown in study. No patients were asked to advise on interpretation online supplementary table 1. or writing up of results. There are no plans to dissemi- Of note, in the beginning of the study period in 2004, nate the results of the research to study participants or notably well before the initiation of the crisis, IMR and the relevant patient community. its components were twofold to fourfold higher among http://bmjopen.bmj.com/

Table 1 Live births, infant deaths, total infant mortality rates (IMR) per 1000 live births in infants born to Greek and non-Greek mothers, and Human Development Index (HDI), in Greece during the period 2004–2016 Infants born to Greek mothers Infants born to non-Greek mothers Mortality Mortality Live Infant rates/1000 Live Infant rates/1000 Year births (N) deaths (N) live births births (N) deaths (N) live births HDI

2004 88 805 403 4.5 16 825 26 1.5 0.835 on October 2, 2021 by guest. Protected copyright. 2005 89 819 383 4.3 17 678 26 1.5 0.845 2006 92 590 379 4.1 19 396 36 1.9 0.851 2007 91 462 359 3.9 20 412 38 1.9 0.849 2008 96 329 281 2.9 21 931 32 1.5 0.857 2009 95 640 307 3.2 22 244 58 2.6 0.858 2010 93 209 355 3.8 21 351 80 3.7 0.856 2011 87 445 282 3.2 18 680 74 4.0 0.852 2012 84 868 241 2.8 15 153 52 3.4 0.854 2013 80 938 292 3.6 13 063 53 4.1 0.856 2014 79 985 298 3.7 12 144 47 3.9 0.864 2015 79 919 320 4.0 11 919 44 3.7 0.866 2016 80 166 338 4.2 12 721 49 3.9 0.868

Excluding missing maternal nationality values.

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 3 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from annually, respectively, p=0.02 for each), whereas no significant change in ENM rate (ENMR) and LNM rate (LNMR) was observed (table 2). Interestingly, mortality trends changed differentially by place of residence; significant decline of IMR trends (−2.1%; −2.9% to −1.3% annually, p<0.0001) and all its components was observed among infants born to Greek mothers living in urban areas, whereas among infants born to Greek mothers in rural areas, a significant increase of IMR (+10.6%; 7.6% to 13.6% annually, p=0.002), ENMR (+24.1%; 17.0% to 31.5% annually, p<0.0001) and LNMR trend (+12.7%; 6.5% to 19.2% annually, p<0.0001) was observed (table 2). By contrast, among infants of non-Greek mothers, the low starting IMR increased significantly (+9.4%; 6.9% to 11.9% annually, p<0.0001) due to significant increases of ENMR, LNMR and PNMR (max ENMR: +12.5%; 8.6% to 16.5% annually, p<0.0001) leading to a non-significant pattern of IMR evolution for Greece overall (p=0.50) (table 2). Further analysis by nationality of non-Greek mothers showed significant increases of IMR trends in all ethnic groups with average annual percent change 7.6% (95% CI 4.2% to 11.1%) in infants born to Alba- nian mothers, 8.6% (3.7% to 13.8%) in infants born to mothers from Balkan countries and countries of the former Soviet Union, 9.4% (2.8% to 16.3%) in infants born to mothers from Asia, 9.4% (1.1% to 18.3%) in infants born to mothers from countries of the European Union or other developed countries and 21.1% (11.0% to 32.3%) in infants born to mothers from Africa. HDI trends increased significantly, by 0.1% (95% CI 0.06% to 0.13%) annually, during the study period. Inverse associations of HDI with IMR, ENMR and PNMR, restricted to Greek mothers’ infants, were observed (table 2). http://bmjopen.bmj.com/ In order to examine whether IMR, in our study population, might have been influenced by any variation in the registration of births in cases of uncertain viability, we analysed separately the infant deaths in the first day of life; trends of IMR in this age group, over the study period, were non-significant (average annual percent change: −1.22%; 95% CI −3.79% to 1.42%; p=0.36). By Joinpoint regression analyses, a break was identified on October 2, 2021 by guest. Protected copyright. in the IMR curve among infants born to Greek mothers restricting the decreasing IMR trend to the period 2004– 2012 (−4.5%; 95% CI −7.6% to −1.3% annually, p=0.01); Figure 1 Evolution of infant early, late and post neonatal a significant decline in ENMR trend until 2011 (−6.5%; mortality rates in Greece (2004–2016) overall and by maternal 95% CI −11.4% to −1.4% annually, p=0.02) was also found nationality. MR, mortality rate/s. (table 3). Of note, however, in the most recent study period following the crisis, non-significant increasing infants born to Greek mothers in comparison with rates trends of IMR (2012–2016: +9.3% annually, p=0.07) in infants born to non-Greek origin mothers. During and ENMR (2011–2016: +10.2% annually, p=0.06) were the subsequent 4–5 years, however, rates among Greeks observed. By contrast, among infants born to non-Greek followed downward trends, whereas among non-Greeks mothers, the high IMR increases (+17.1%; 95% CI 8.1% inflated reaching almost similar values to those observed to 26.9% annually, p=0.002) were restricted to the period in infants born to Greek mothers (figure 1). 2004–2011 with no statistically significant change there- Overall, during the study period, among infants born after (2011–2016); no break was identified for ENMR to Greek mothers, Poisson regression analyses showed curve. Significant increases in IMR and ENMR trends decreasing trends in IMR and PNM rate (PNMR) (−0.9%; after the year 2008 were observed in the total study popu- 95% CI −1.7% to −0.1% and −1.6%; −3.0% to −0.2% lation, whereas no break was found for LNMR and PNMR

4 Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from Pvalue CI)

APC (95% PC −12.3 (−22.2 to –1.2) 0.03 −1.3 (−30.7 to 40.5) 0.94 −11.4 (−20.8 to –0.7) 0.04 PNMR P value N CI)

3.5 (−10.9 to 20.1) 0.66 3.3 (−10.4 to 19.3) 0.65 20.0 (−28.6 to 101.7) 0.49 APC (95% PC LNMR P value N CI)

http://bmjopen.bmj.com/ APC (95% PC −11.3 (−20.8 to –1.4) 0.03 −19.8 (−44.8 to 16.6) 0.25 −14.0 (−22.3 to –4.7) 0.004 ENMR on October 2, 2021 by guest. Protected copyright. P value N CI)

APC (95% PC −8.6 (−14.8 to −1.9) 0.01 −5.7 (−25.0 to 18.7) 0.62 −9.5 (−15.3 to −3.2) 0.004 IMR N 4238615 −0.9 (−1.7 to −0.1) 0.024862 9.4 (6.9 to 11.9) 0.3 (−0.5 to 1.0) 1849 <0.0001 0.50 2583850 −0.6 (−1.8 to 0.7)563 −2.1 (−2.9 to −1.3) 0.35 2107 12.5 (8.6 to 16.5) <0.00014421 8.8 (6.3 to 11.4) 1738 <0.0001 0.9 (−0.3 to 2.1) 995 −0.8 (−1.6 to 0.0) 138 <0.0001388 0.05 −2.0 (−3.3 to –0.8) 0.13 −0.6 (−2.3 to 1.1) 243 8.0 (3.1 to 13.2)38 0.002 0.46 10.6 (7.6 to 13.6)427 1981 1136 0.001 906 0.002 13.2 (9.2 to 17.4) 14.0 (4.0 to 24.9) 0.4 (−1.2 to 2.0) 1393 10.9 (8.1 to 13.9) −0.3 (−1.5 to 0.9) 218 <0.0001 −1.8 (−3.5 to 0.0) 0.005 108 123 <0.0001 0.65 −1.6 (−3.0 to –0.2) 0.61 0.05 118 10 4.4 (0.6 to 8.4) 0.02 6.7 (1.5 to 12.2) 24.1 (17.0 to 31.5) 1617 1037 1206 <0.0001 24.3 (17.6 to 31.4) 0.02 0.01 −0.8 (−2.4 to 0.9) 93 15.5 (−2.4 to 36.8) −0.9 (−2.2 to 0.5) −2.4 (−3.9 to –0.9) <0.0001 0.09 0.36 104 0.20 0.002 197 12.7 (6.5 to 19.2) 14.7 (8.6 to 21.1) 11 1408 <0.0001 4.9 (0.9 to 9.2) 187 <0.0001 −1.5 (−2.9 to 0.0) 18.1 (0.3 to 39.0) 205 0.02 3.3 (−0.6 to 7.4) 0.04 0.05 3.0 (−0.7 to 6.9) 0.09 17 0.11 4.9 (−7.6 to 19.1) 0.46 egression derived changes and 95% CI of infant (IMR), early (ENMR), late (LNMR) and post neonatal mortality rates (PNMR): annual percent of change derived changes and 95% CI of infant (IMR), early (ENMR), late (LNMR) post neonatal mortality rates (PNMR): annual percent egression Poisson r

* * Greek Non-Greek Any Greek Non-Greek Any * Greek Non-Greek All* Urban–semiurban Rural (APC) for place of residence and maternal nationality (2004–2016); percept of change (PC) for 1 SD of Human Development Index (HDI) controlling for year of birth of change (PC) for 1 SD Human Development Index (HDI) controlling and maternal nationality (2004–2016); percept (APC) for place of residence Table 2 Table

Maternal nationality Urbanisation HDI *Including missing maternal nationality values. Greek Non-Greek All

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 5 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from measures in Greeks, non-Greeks or the total study population (table 3). The results of ITS analyses in the preset years 2008, 2010 and 2012 are shown in figures 2 and 3. Compared with the joinpoint regression showing non-significant P value P value increases in IMR from 2012 to 2016 for infants born to Greek mothers, ITS analysis showed a significant increase in IMR (p=0.0001) in this population during the same time period and further identified significant increases

CI) CI) in LNMR (p=0.0004) and PNMR trends (p=0.03).

Regarding the non-significant increases of ENMR trends after the year 2011 by Joinpoint regression in infants born to Greek mothers (p=0.06), ITS analysis showed a

APC (95% APC (95% non-significant increase during the post 2012 time period (p=0.07); however, a significant increase during the post 2010 time period in these infants was observed (p=0.01) (figure 2). In infants born to non-Greek mothers, while a

218 4.7 (−0.8 to 10.6)significant 0.09 increase in IMR trends was found by Joinpoint 13931617 −1.7 (−4.0 to 0.8) −0.8 (−3.0 to 1.4) 0.16 0.42 regression analyses during the period 2004–2011, by ITS analyses—depending on when the year of interest was being defined (2010 or 2012)—the increase was either not significant (2004–2010, p=0.11) or significant (2004– 2012, p=0.001); no change in IMR was identified by ITS analysis over the 2010–2016 or 2012–2016 time periods in

ENMR PNMR non-Greek mothers’ infants (figure 3).

Discussion Trends of neonatal and infant mortality in Greece, P value Time period N P value Time period N 0.02 2008–2016 1327 6.1 (2.1 to 10.3)spanning 0.01 from 2004—before the initiation of crisis in 2008—to 2016, are not homogeneous, given significant increases in the relatively small proportion of infants born to non-Greek nationality mothers as contrasted to http://bmjopen.bmj.com/ declining IMR trends among Greek mothers’ infants, CI) CI) differentials by place of residence and also increases in IM indices among Greek mothers’ infants noted ~4 years after the crisis initiation. Specifically, the ‘brake’ of 2012

9.1 (3.5 to 14.9)0.6 (−1.3 to 2.5) 0.003 0.48identified, 2004–16 2004–16 by Jointpoint regression analysis, non-sig- APC (95% APC (95% −0.6 (−2.5 to 1.4) 0.54 2004–16 −4.5 (−7.6 to −1.3) 0.01 2004–2011 1118 −6.5 (−11.4, to −1.4) 0.02 9.3 (−1.1 to 20.9) 0.07 2011–2016 731 10.2 (−0.4 to 21.9) 0.06 17.1 (8.1 to 26.9) 0.002 2004–2016 258 13.8 (6.7 to 21.3) 0.001 0.2 (−11.5 to 13.5) 0.97 −7.5 (−14.6 to 0.3) 0.06 2004–2008 780 −10.7 (−19.9 to −0.5) 0.04 3.6 (0.7 to 6.6) nificant increases of ~10% annually, in IMR and ENMR trends, whereas by ITS analyses, significant increases in IMR, LNMR and PNMR trends were observed thereafter on October 2, 2021 by guest. Protected copyright. among infants born to Greek mothers, a fact indicating how sensitive the results are depending on the statistical method used. Changes in HDI during the study period 138 995 were also reflected in the inverse associations of HDI with ENMR and IMR, but also PNMR, among infants born to Greek mothers. Strengths of the present investigation, as contrasted to previously Greek-based published studies,9 18 20 21 23 include the longest follow-up period of national data, egression derived annual percent of change (APC) and 95% CIs infant (IMR), early neonatal (ENMR), late (LNMR) post derived annual percent egression IMR Time period N LNMR Time period N 2004–16 1136 2004–2012 2870 2004–16 2012–2016 1368 2011–2016 282 2004–2008 1807 2008–2016 3055 namely since 2004, before the crisis, when the first data on maternal ethnicity became electronically available, to the latest available year 2016. Of great importance is also the fact that, for the first time, the maternal ethnicity was Joinpoint r

considered in the analyses explaining part, if not most, of the heterogeneity of trends through the study period.

* To this end, three complimentary statistical methodolo- Table 3 Table (PNMR) mortality rates by maternal nationality No break in the time period examined was identified for ENMR in non-Greeks, as well as for LNMR and PNMR measures in Greeks, non-Greeks or the total study population. non-Greeks in Greeks, as well for LNMR and PNMR measures in the time period examined was identified for ENMR non-Greeks, No break *Including missing maternal nationality values. All Non-Greek 2004–16 Greek Greek Non-Greek 2004–2011 333 All* gies were used to describe the overall evolution depicted

6 Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from http://bmjopen.bmj.com/ on October 2, 2021 by guest. Protected copyright.

Figure 2 Interrupted time series analyses for evolution of infant (IMR), early (ENMR), late (LNMR) and post neonatal mortality rates (PNMR) in alternative prespecified years of interest (2008, 2010 and 2012) in infants born to Greek mothers. in figure 1. Thus, we were able to enhance the validity based on the number of live births; yet, the absence of a of the results pertaining to the study of the evolution of linkage system between birth and death data consists a individual IMR components, namely ENMR, LNMR and limitation of the study. The absence of a linkage system, PNMR, in association with socioeconomic factors (ie, as well as that of other official registries in Greece, has maternal nationality, place of residence and HDI) for not allowed use of known determinants of socioeconomic which individual data have been available on a nation- deprivation or social coherence, such as parental educa- wide level among both deaths and live births during the tion, employment/occupation, marital status, family study period. Calculation of IMR and its components was income and household size, neither were data on the role

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 7 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from http://bmjopen.bmj.com/ on October 2, 2021 by guest. Protected copyright.

Figure 3 Interrupted time series analyses for evolution of infant (IMR), early (ENMR), late (LNMR) and post neonatal mortality rates (PNMR) in alternative prespecified years of interest (2008, 2010 and 2012) in infants born to non-Greek mothers. of biological factors, such as parental age, gestational age factors.25 Lastly, analyses by specific cause of infant death and birth weight, or multiplicity of pregnancy available were not set among the outcome measures of this study. for both the live births and deaths series.5 19 24 Besides, Assessing migration as a determinant of perinatal health the study period, with available information on maternal outcome and infant death in the developed countries is nationality and place of residence, was rather short a complex undertaking resulting in conflicting results of leading to ITS analyses with few data points; however, published studies. Specifically, some studies have shown ITS analysis can be used even for few observations; no that ethnic minority is a significant risk factor for unfa- log transformation is needed but the results should be vourable perinatal outcomes and increased neonatal interpreted with caution as the power depends on various and infant mortality,18 26–33 especially if it coincides with a

8 Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from financial crisis, as was the case in Greece. Factors possibly contribute to increases in IMR among Greek children contributing to the increased risk for poorer outcomes following the crisis. among children born to migrants include socioeconomic Indeed, the strongly decreasing IMR trends among disadvantage, poor communication, discrimination, Greek infants were stagnated after a lag time of ~4 reduced utilisation of health facilities, low quality of care, years of crisis and equalised with the previously sharply and also stress and consanguinity or differing attitudes to increasing trends among non-Greek maternal nation- screening and termination of pregnancy associated with ality infants; this time lag is reasonable43 and in line with preterm birth, low birth weight or congenital anomalies previous findings showing that following reductions in and lethal inherited diseases.26 30 34–36 By contrast, other the government healthcare spending in Europe, the studies present similar or even better perinatal health greatest negative effect on neonatal and post neonatal outcomes among some migrant groups than among mortality was 4 and 5 years later, respectively.6 The irony natives (‘healthy migrant effect/paradox’).26–28 37 38 Refu- is that the period 2012–2016, during which the crisis was gees have been recognised as the most vulnerable group deepening13 14 and stricter economic austerity measures suffering increased severe neonatal morbidity and infant were applied,44 disparities gap in IMR trends diminished mortality risk,26 28 although no absolutely clear pattern against the Greek population; IMR trends deteriorated regarding refugee or non-refugee status among migrants only in infants born to Greek mothers, whereas in infants has been identified.34 born to non-Greek mothers, IMR trends approached an Ιn our study, starting rates of IMR and its compo- anelastic highest value up to 2011 with no statistically nents among infants born to non-Greek mothers were significant change thereafter. lower compared with those of Greek mothers. Of note, Disparities in IMR exist across geographic areas even the ratio of non-Greek infants among live births during in well-developed countries; that is, IMR in the USA vary 2009–2016 versus 2004–2008 was only 1.32, whereas by urbanisation level of maternal residency being lowest the respective figure among infant deaths was almost in large urban counties but highest in rural areas.45 threefold (data not shown). Actually, in the precrisis Rural women may face health challenges related to era, non-Greek mothers of newborns comprised mainly geographic barriers to care (less timely and/or appro- economic migrants, most of them of Albanian nation- priate care) and physician shortages, but they may also ality living for many years, or even born, in Greece. After present differences in a number of socioeconomic and 2008, Greece experienced an influx of refugees with the demographic risk factors, such as less education, lower majority of them fleeing from war and terror in Syria, income, younger age at pregnancy or greater number Occupied Palestinian Territory, but also from Afghani- of children, in comparison with their urban counter- stan, Iraq and other countries.16 17 It could be suggested parts.46 47 The increasing IMR trends during the study that changes in the homogeneity of foreign mothers period in infants born to Greek mothers living in rural during the study period, in association with the country’s areas, as opposed to the declining IMR trends in those http://bmjopen.bmj.com/ economic difficulties, might have led from the ‘healthy born to Greek mothers living in urban/semiurban areas, migrant effect’, observed before 2008, to the deterio- is worrisome. Specific causes should be further studied ration of IM indicators and IMR trends, afterwards, in and addressed. Regarding infants born to non-Greek infants born to non-Greek mothers. We did observe mothers, the increasing IMR trends despite living in urban significant increases in IMR trends in all ethnic groups of areas can be explained by previous reports showing that non-Greek mothers during the study period, especially other risk factors, that is, young maternal age (<20 years) in infants born to mothers from Africa; however, as the or maternal ethnic minority, may be more powerful than numbers by individual maternal nationality were rather place of residence.47 on October 2, 2021 by guest. Protected copyright. small, this issue needs to be further studied. Among the three IMR components, ENMR seemed to Simultaneously with immigration to/through Greece, be the most ‘sensitive’ in reflecting adverse impacts on the country experienced the third wave of mass emigra- child health. Specifically, ENMR was overall positively tion in the 20th and 21st centuries; over 400 000 Greek associated with rural place of residence of Greek mothers' citizens left Greece since the onset of the economic crisis, children and also presented the most sizeable increase in 2008, seeking new opportunities and employment in among IM indices in non-Greeks (table 2). Moreover, other countries, mainly Germany, the UK and the Neth- among the three IM components, ENMR was the first or erlands.39 40 The current emigration wave of Greeks the only one, depending on the statistical method used, involved mostly highly educated people (the so called that increased after the crisis initiation in infants born to ‘brain-drain’) of young age, leading to a decrease in the Greek mothers. ENMR, in developed countries, including number of Greek women of childbearing age.41 This fact, Greece, represents more than 70% of neonatal deaths along with the decrease in the fertility rate by almost on account mainly of prematurity/low birth weight and 10% between 2008 and 2016, contributes significantly congenital anomalies.14 48–51 Actually, it reflects perinatal to the reduction in the annual number of live births in health and care during pregnancy and labour, and also Greece by almost 23% during the period of the economic postnatal care in the first week of life.50 Increased inci- crisis.41 42 Furthermore, the flow of mostly affluent and dence of impaired perinatal parameters including low well-educated people going outside Greece could also birth weight, prematurity and increased maternal age

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 9 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from and rate of caesarian section has been reported during Author affiliations 1 the years of the economic decline in Greece.18 19 Cuts First Department of Pediatrics, School of Medicine, National and Kapodistrian University of Athens, Athens, Greece in public health expenditures between 2008 and 2016, 2Department of Hygiene, Epidemiology and Medical Statistics, School of Medicine, reduction in healthcare workforce and paediatric nurses National and Kapodistrian University of Athens, Athens, Greece as well as reduction in the number of obstetrics beds, 3Department of Hygiene and Epidemiology, University of Ioannina Medical School, obstetricians and midwifes (−45.5%, −60.2% and −27.5%, Ioannina, Greece 14 4Department of Epidemiology and Biostatistics, Imperial College, London, UK respectively) could possibly explain, at least in part, 5 the observed positive ENMR trends during the postcrisis Clinical Epidemiology Unit, Karolinska Institute, Stockholm, Sweden period in Greece. Acknowledgements ELSTAT has provided individual anonymised data. The exact mechanisms leading to the disparities in Contributors EP and GC were involved in the initial conception of the study. All trends of neonatal and infant mortality observed in this authors contributed to the design, analysis and interpretation of data. TS, ND and study remain to be further explored. Specific maternal, EP developed the first draft of the manuscript and all authors critically revised it and family and infant features, including detailed and punc- approved the final version. ND, AA and CM performed the statistical analyses. EE tual information on the causes of infant deaths, would reviewed statistics and contributed significantly for ITS analyses. All authors agree to be accountable for all aspects of the work in ensuring that questions related to have shed more light on the links between IMR trends and the accuracy or integrity of any part of the work are appropriately investigated and socioeconomic factors but were beyond the scope of this resolved. The corresponding author ensures that all the journal’s administrative article and left for future research. Meanwhile, policies and requirements are properly completed. programmes should be implemented to mitigate the nega- Funding The authors have not declared a specific grant for this research from any tive impact of the crisis on population and infant health in funding agency in the public, commercial or not-for-profit sectors. Greece. Vulnerable groups, such as mothers of non-Greek Competing interests None declared. nationality, of low income or of rural place of residence, Patient consent for publication Not required. should be specifically addressed and their rights to health Provenance and peer review Not commissioned; externally peer reviewed. protected. Barriers in the access of refugees to the Greek Data sharing statement Data are available upon request from ELSTAT (http:// healthcare system have been identified; they are mostly www.statistics. gr/el/ statistical-data- request) related to language, culture and inadequate information Open access This is an open access article distributed in accordance with the about the healthcare system, but also include difficulties in Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which the coordination of Health Services, transportation prob- permits others to distribute, remix, adapt, build upon this work non-commercially, lems, issues in obtaining expertise medical assessment in and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use the camps, lack of continuity of care, financial difficulties is non-commercial. See: http://creativecommons.org/ licenses/by- nc/4. 0/. in making out-of-pocket payments for health and social care services and administrative barriers, among others.52 In response, the Greek government and several non-governmental organisations initiated commendable actions, http://bmjopen.bmj.com/ that is, National Health System services free of charge for References 1. Petridou E, Valadian I, Trichopoulos D, et al. Medical services and uninsured and vulnerable social groups including asylum socioeconomic factors: Determinants of infant mortality in Greece. seekers, translation services in public hospitals, access Hygie 1989;3:20–3. to specialist care/treatment with gynecologists (mostly 2. WHO. The determinants of health. http://www. who.int/hia/evidence/ doh/en (Accessed 30/5/2018). women), midwives, dentists, psychologists and psychia- 3. Zylbersztejn A, Gilbert R, Hjern A, et al. Child mortality in trists being lately included in the camp clinics.52 These England compared with Sweden: a birth cohort study. Lancet 2018;391:2008–18. actions may have prevented deterioration of IMR which 4. Rajmil L, Fernandez de Sanmamed MJ, Choonara I, et al. are being kept at steady levels after 2011–2012 in infants International Network for Research in Inequalities in Child Health (INRICH). Impact of the 2008 economic and financial crisis on on October 2, 2021 by guest. Protected copyright. of non-Greek mothers. Efforts should be continued and child health: a systematic review. Int J Environ Res Public Health intensified, however, to ensure equity with local popula- 2014;11:6528–46. tions. Strong governmental integration policy for minori- 5. Glonti K, Gordeev VS, Goryakin Y, et al. A systematic review on health resilience to economic crises. PLoS One 2015;10:e0123117. ties paired with initiatives to improve social coherence, a 6. Budhdeo S, Watkins J, Atun R, et al. Changes in government deeply rooted mechanism for protection of health among spending on healthcare and population mortality in the European union, 1995–2010: a cross-sectional ecological study. J R Soc Med those in need in the Greek society, along with further 2015;108:490–8. improvements in primary healthcare delivery could help 7. Tapia Granados JA, Rodriguez JM. Health, economic crisis, and to alleviate the cost in infant lives and ensure healthy austerity: A comparison of Greece, Finland and Iceland. Health Policy 32 51 2015;119:941–53. adulthood. Irrespective of the crisis, improvements 8. Regidor E, Barrio G, Bravo MJ, et al. Has health in Spain been in the quality of perinatal and neonatal care, including declining since the economic crisis? J Epidemiol Community Health 2014;68:280–2. centralisation of very preterm deliveries, establishment of 9. Laliotis I, Ioannidis JPA, Stavropoulou C. Total and cause-specific regional perinatal centres, monitoring of the implementa- mortality before and after the onset of the Greek economic tion of evidence-based practices in maternity and neonatal crisis: an interrupted time-series analysis. Lancet Public Health 2016;1:e56–e65. units, as well as increase in health expenditures, health- 10. Parmar D, Stavropoulou C, Ioannidis JP. Health outcomes during care workforce, number of doctors and midwifes/nurses, the 2008 financial crisis in Europe: systematic literature review. BMJ 2016;354:i4588. could decrease ENMR, eventually leading to overall 11. O’Donnell C, Burns N, Dowrick C, et al. Health-care access for 50 decrease in IMR in Greece. migrants in Europe. Lancet 2013;382:393.

10 Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 Open access BMJ Open: first published as 10.1136/bmjopen-2018-025287 on 18 August 2019. Downloaded from 12. Karanikolos M, Heino P, McKee M, et al. Effects of the Global 32. Bollini P, Pampallona S, Wanner P, et al. Pregnancy outcome of Financial Crisis on Health in High-Income Oecd Countries: A migrant women and integration policy: a systematic review of the Narrative Review. Int J Health Serv 2016;46:208–40. international literature. Soc Sci Med 2009;68:452–61. 13. Organisation for Economic Co-operation and Development OECD 33. Gotsens M, Malmusi D, Villarroel N, et al. Health inequality data. 2016. https://data.oecd.org [Accessed 30 May 2018]. between immigrants and natives in Spain: the loss of the healthy 14. Hellenic Statistical Authority Statistics. 2016. www.statistics. gr immigrant effect in times of economic crisis. Eur J Public Health [Accessed 30 May 2018]. 2015;25:923–9. 15. Zavras D, Zavras AI, Kyriopoulos II, et al. Economic crisis, austerity 34. Nybo Andersen AM, Gundlund A, Villadsen SF. Stillbirth and and unmet healthcare needs: the case of Greece. BMC Health Serv congenital anomalies in migrants in Europe. Best Pract Res Clin Res 2016;16:309. Obstet Gynaecol 2016;32:50–9. 16. European Union. The EU and the migration crisis. 2017. http:// 35. Di Renzo GC, Tosto V, Giardina I. The biological basis and publications.europa.eu prevention of preterm birth. Best Pract Res Clin Obstet Gynaecol 17. Mertens E, Rockenschaub G, Economopoulou A, et al. Assessment 2018;52:13–22. of public health issues of migrants at the Greek-Turkish border, April 36. Wanigaratne S, Cole DC, Bassil K, et al. The influence of refugee 2011. Euro Surveill 2012;17:pii: 20056. status and secondary migration on preterm birth. J Epidemiol 18. Zografaki I, Papamichail D, Panagiotopoulos T. Adverse effect of the Community Health 2016;70:622–8. financial crisis in Greece on perinatal factors. Eur J Public Health 37. Kirby RS. Perinatal outcomes and nativity: does place of birth really 2018;28:1116–21. influence infant health? Birth 2011;38:354–6. 19. Sdona E, Papamichail D, Ragkou E, et al. Greek economic crisis and 38. Sdona E, Papamichail D, Ragkou E, et al. Migration status and impaired perinatal parameters: experience from a public maternity perinatal parameters in a Greek public maternity hospital: an hospital. J Matern Fetal Neonatal Med 2018;31:2371–5. illustration of the "healthy immigrant effect". J Matern Fetal Neonatal 20. Michas G, Varytimiadi A, Chasiotis I, et al. Maternal and child Med 2019;32:62-6. mortality in Greece. Lancet 2014;383:691–2. 39. Eurostat. Emigration by age group, sex and citizenship. http:// 21. Filippidis FT, Gerovasili V, Millett C, et al. Medium-term impact of the appsso.eurostat.ec.europa.eu/nui/show.do?dataset=migr_emi1ctz& economic crisis on mortality, health-related behaviours and access to lang=en [Accessed 04/04/2018]. healthcare in Greece. Sci Rep 2017;7:46423. 40. Cavounidis J. The changing face of emigration: Harnessing the 22. Kim HJ, Fay MP, Feuer EJ, et al. Permutation tests for joinpoint potential of the New Greek Diaspora. Washington, DC: Migration regression with applications to cancer rates. Stat Med 2000;19:335–51. Policy Institute, 2015. 23. Simou E, Koutsogeorgou E. Effects of the economic crisis on health 41. Vrachnis N, Vlachadis N, Iliodromiti Z, et al. Greece's birth rates and and healthcare in Greece in the literature from 2009 to 2013: a the economic crisis. Lancet 2014;383:692–3. systematic review. Health Policy 2014;115:111–9. 42. Eurostat, “Fertility indicators”. http://appsso.eurostat.ec.europa.eu/ 24. Petridou ET, Georgakis MK, Erdmann F, et al. Advanced parental age nui/show.do?dataset=demo_find&lang=en [Accessed 20/09/2018]. as risk factor for childhood acute lymphoblastic leukemia: results 43. Karanikolos M, Mladovsky P, Cylus J, et al. Financial crisis, austerity, from studies of the Childhood Leukemia International Consortium. Eur J Epidemiol 2018;33:965–76. and health in Europe. Lancet 2013;381:1323–31. 25. Bernal JL, Cummins S, Gasparrini A. Interrupted time series 44. Burki T. Health under austerity in Greece. Lancet 2018;391:525–6. regression for the evaluation of public health interventions: a tutorial. 45. Ely DM, Driscoll AK, Mathews TJ. Infant mortality rates in rural and Int J Epidemiol 2017;46:348–55. urban areas in the United States, 2014. NCHS data brief, No 285. 26. Gissler M, Alexander S, MacFarlane A, et al. Stillbirths and infant Hyattsville, MD: National Center for Health Statistics, 2017. deaths among migrants in industrialized countries. Acta Obstet 46. Van de Poel E, O'Donnell O, Van Doorslaer E. What explains Gynecol Scand 2009;88:134–48. the rural-urban gap in infant mortality: household or community 27. Gagnon AJ, Zimbeck M, Zeitlin J, et al. Migration to western characteristics? Demography 2009;46:827–50. industrialised countries and perinatal health: a systematic review. 47. Harris DE, Aboueissa AM, Baugh N, et al. Impact of rurality on Soc Sci Med 2009;69:934–46. maternal and infant health indicators and outcomes in Maine. Rural 28. Wanigaratne S, Cole DC, Bassil K, et al. Severe Neonatal Remote Health 2015;15:3278. Morbidity Among Births to Refugee Women. Matern Child Health J 48. Heron M. Deaths: Leading Causes for 2015. Natl Vital Stat Rep

2016;20:2189–98. 2017;66:1–76. http://bmjopen.bmj.com/ 29. Puthussery S. Perinatal outcomes among migrant mothers in the 49. Roncancio CP, Misnaza SP, Peña IC, et al. Trends and characteristics United Kingdom: Is it a matter of biology, behaviour, policy, social of fetal and neonatal mortality due to congenital anomalies, Colombia determinants or access to health care? Best Pract Res Clin Obstet 1999-2008. J Matern Fetal Neonatal Med 2018;31:1748–55. Gynaecol 2016;32:39–49. 50. Lehtonen L, Gimeno A, Parra-Llorca A, et al. Early neonatal death: A 30. Wanner P, Bollini P. The contribution of the foreign population to the challenge worldwide. Semin Fetal Neonatal Med 2017;22:153–60. high level of infant mortality in Switzerland: a demographic analysis. 51. Alexander M, Harding M, Lamarche C. Quantifying the impact of BMC Pregnancy Childbirth 2017;17:151. economic crises on infant mortality in advanced economies. Appl 31. Cantarutti A, Franchi M, Monzio Compagnoni M, et al. Mother's Econ 2011;43:3313–23. education and the risk of several neonatal outcomes: an evidence 52. Kotsiou OS, Kotsios P, Srivastava DS, et al. Impact of the Refugee from an Italian population-based study. BMC Pregnancy Childbirth Crisis on the Greek Healthcare System: A Long Road to Ithaca. Int J

2017;17:221. Environ Res Public Health 2018;15:1790. on October 2, 2021 by guest. Protected copyright.

Siahanidou T, et al. BMJ Open 2019;9:e025287. doi:10.1136/bmjopen-2018-025287 11